Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Processors/DSPs

Kontron packet processor blade offers future proofing

Posted: 17 Feb 2011 ?? ?Print Version ?Bookmark and Share

Keywords:ATCA? packet processor blade? telecom? hardware?

Kontron AG debuts the AT8242 40GbE ATCA packet processor blade with two OCTEON II cnMIPS64 multicore processors from Cavium Networks.

The AT8242 manages complex networking functions and provides a future-proof investment that will be able to scale with the expected exponential growth of converged voice, video and data traffic over IP-based networks.

It features two OCTEON II CN6880 processors, each one integrating 32 enhanced cnMIPS64 v2 cores with up to 48GHz of 64bit computing power in a single chip. It also combines over 85 L3-L7 application acceleration engines, virtualization features, 100Gbit/s of connectivity, and a new Real Time Power Optimizer that dynamically adjusts power depending upon the application-level processing requirement.

Customers can use the Kontron 40GbE ATCA packet processor blade AT8242 in combination with the Kontron 14-slot ATCA platforms OM9141-40G or OM9141-10G, featuring switch management software and up to four times the I/O switching bandwidth of previous platforms.

The Kontron AT8242 is designed for versatility and can be used in many core and edge network applications, particularly Packet Data Network Gateways (PGW), Serving Gateways (SGW) and Mobility Management Entity (MME) network elements found in the 4G LTE Evolved Packet Core (EPC). Growth in mobile data traffic networks will require more efficient and powerful deterministic processing of applications such as packet processing, deep packet content inspection (RegEx), application-aware switching and routing, parsing, IPv6/IPv4 forwarding, compression/decompression, classification, and QoS at 40Gbit/s wire-speed.

Enhanced 64bit cores
Cavium Networks has redesigned its cores from the previous generation to offer a better TCP and control plane performance per core with the OCTEON II multicore processors. Aside from enhanced MIPS cores, the processors also feature dedicated hardware blocks, more memory, cache-set association up to 37-way and a variety of new interfaces.

Each of the two OCTEON II processors C linked via Interlaken x4 interfaces C provides up to 48GHz of compute performance, and 37KB I-cache and 32KB D-cache operating at up to 1.5GHz for 3 to 4 times more performance than alternative solutions. The resulting processing and interface architecture means the CN6880 is capable of 40Gbit/s duplex line rate processing, 20Gbit/s data compression and its inbuilt DPI engines can handle 40Gbit/s of application recognition independent of the number of rules and number of flows.

DPI functions are handled by Cavium Networks' Hyper-Finite Automata (HFA) engines, a new innovative hardware acceleration block that, in addition to the packet header information, also examines the footer, source, destination and payload information to help identify applications and protocols for more efficient network management of services. Another vital on-chip processor engine is the application acceleration manager, which processes incoming packets, queues and tags them, schedules a task to an available core and decides what will happen to the result.

The CN6880 supports 4MB of partitionable L2 cache using a HyperConnect crOS-bar with more than 6Tbit/s throughput. This means lower latency and improved I/O traffic performance. In addition, each cnMIPS64 core has a dedicated 37KB L1 l-cache and 32KB L1 D-cache, which provides deterministic, lower latency and improved I/O traffic performance and avoids cache 'thrashing'.

1???2?Next Page?Last Page

Article Comments - Kontron packet processor blade offer...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top